Hi all, conveyor merge section, 1734-AENTR with a long rack of IB8s and OB8s hanging off it. The modules past about slot 8 or 9 fault and come back on their own, yellow triangle for a few seconds then green. Worse when the section runs flat out with most of the outputs on. Slots 1 through 8 never fault, and the AENTR itself never faults either, both link lights fine.
We swapped the terminal base under slot 12, the one that drops the most, and slowed the RPI on the far modules from 10ms to 50ms. Neither made any difference. It's been like this on and off for months and it's getting worse. Is the AENTR dying, or am I missing something obvious?
How many modules exactly, what's in slots 1 to 20 in order, and is there a 1734-EP24DC anywhere in the rack? What's the AENTR firmware? A 20 module POINTBus without an expansion power module is where I'd look before the adapter.
20 exactly. No EP24DC, never heard of it. Firmware on the AENTR is 5 something. Order is 8 IB8, 4 OB8, 2 OE2V, then 6 more OB8 on the end. The OB8s at the end are the ones on the most when it faults.
Seen it on a sorter with 18 modules. The AENTR pushes about 1A down the POINTBus for the modules' own electronics and 20 modules with output cards eat that up. Far end browns out under load, drops off the bus, comes back when it recovers. That's your self healing triangle.
What you want is a 1734-EP24DC. It's a module, goes in the rack where the budget runs out, starts a fresh POINTBus segment for everything after it. Counts as a slot in the tree, so slot numbers shift, and it needs its own 24V. Add up the POINTBus mA per module from 1734-SG001, put it after the last one the adapter can carry, probably slot 9 or 10.
Ordered one, be here Thursday. Adding up from the SG001 I get about 1.05A for all 20, which is right at the edge, and the OB8s at the end are where it tips. Makes sense with what I see. Slot shift is going to be a pain, every alias on those last 11 modules.
Put the EP24DC in as slot 9 in the tree before it arrives, offline, and use Find and Replace for the aliases. Ten minutes.
EP24DC in at slot 9 Thursday night, 24V and common on it, tree done offline first with Find and Replace on the aliases like rsdave said. Downloaded, all 21 green. Ran flat out all of Friday, not one triangle.
So it was the POINTBus budget the whole time. Twenty modules on one segment, the adapter's 1A used up, the far end dropping out under load. One EP24DC at slot 9 and everything after it has its own supply now.
What I don't get is how it ran 3 years before it started. Maybe the supply sagged over time, or somebody added the last OB8s and never said. Probably that.
Thanks carlh and rsdave.
Marking solved.
plctr.com team